Reverb insertion losses

I'm having problems designing a decent reverb for guitar.

I really don't like the way the fender 12AT7/12AX7 units sound as I think they have too much reverb. The same circuit also has too much insertion loss to really be practicle without that extra gain stage before the PI.

I think most of the insertion loss is due to the fact that in order to avoid oscilations you can't go below ~2 megohms on the bypass resistor. Unfortunatly this means that when the reverb is off you lose a lot of gain from your preamp stages because the signal still needs to pass through that resistor (and/or xx picofarad cap.)

I think I may have devised a circuit that gets rid of at least the insertion loss of the preamp circit, still working on how to bring the reverb gain up to unity.

[IMGDEAD]http://getchellaudio.googlepages.com/ReverbMix.jpg[/IMGDEAD]

Ideas? Thoughts? Opinions? :)

P.s. : Any decent 1 tube Send/recovery reverb circuits out there, or is there simply not enough gain? Thoughts on capacitvly coupled reverb tanks? IE ~2500 ohm tanks being driven diectly by 1/2 a 12xx7 tube
 
Look at Kevin O'Connors website - London Power.

He has a stunning reverb driver using 12AT7 in push pull. This is what I use.

For low cost options, the Accutronics webpages have a number of solid state reverb driver circuits. The Current drive variation from there will give "tube like" performance.

The large series resistor in the "dry" chain is not the only way of handling the mix of the reverb signal but is the cheapest. Other techniques virtually require you to dedicate another tube stage as a specialist mixer.

If you do a bit of messing with guitar amps it would probably be worth buying one or more of Kevin O'Connor's "The Ultimate Tone" (TUT) series books. TUT 1 includes a section on reverbs.

Cheers,
Ian
 
I looked at the reverb on his site, but the transformer is way overpriced for what it is. Its probably a $10 part from hammond or something. I think that going into a push-pull output stage for a reverb is a bit too involving when trying to cut costs and still have good sound quality.

I have also been looking at cap coupled tanks. Any opinions on this topology?
Also what are opinions of running a high gain pre into a reverb unit... there would be some problems with overdriving the pan wouldn't there, unless you added a stopper resistor?
 

m6tt

Member
2007-08-08 12:12 am
Pan overdrive doesn't sound too bad...at least with a doubling of gain in my hacked amp (an au7 replaced with a beefy ax7)...the original fender standalone used a 6k6 (essentially a 6v6 with lower ratings). I imagine a 6aq5, 6bq5 or might make a nice driver as well. Just be careful where the tank is in relation to speakers, or other vibrations such as 'bouncy' floors if recording especially. I had a hell of a time tracking down a sudden increase in high-frequency gain, but I believe it was actually the speakers rear sound projections manipulating the reverb springs! Also, the circuit you demonstrate is interesting...in this amp, originally a hammond spinet, there is a pot to control reverb, (1meg) where as resistance is increased, reverb also increases...it may be some type of feedback, as it connects at the phase inverter. A similar circuit to yours is used for pre-amp / reverb recovery mixing however. I think its an AO-48, but i've never seen one anywhere, never seen a schematic, just the one I have.
 
Thanks for the info! I believe your right about the feedback. If you don't have some sort of resistor from the beginnning to the end the sound is a lot quieter. I don't really want to use a power tube to drive the tank because I want to keep cost and power consumption ( and therefore cost of power/ another output transformer) down.

if I could find a suitable tank I would eliminate the OPT all together and use a capacitor coupled theme. My reverb is going to be mounted in a head with the amp, but I have located the unit in a small combo before. The only feedback I have ever gotten is from a microphonic tube (usually the driver tube). BTW, I am going to be ordering TUT 1 and probably some others in his series soon.
 

revintage

Member
2007-05-30 4:12 pm
Hi,
To get less loss you have to change the topology a little:

1 Add a DC-coupled CF to the gainstage before the split.

2 Make the dry signal resistor series resistor about 100k.

3 Put an input (dwell) control at the input of the reverbdriver tube.

4 Skip the parallelled 12AT7 and use an EL84(cheap and easy to find) biased to ca 15mA at 300V(no need for beef-uped power supply) together with a standard Fender reverb transformer.

5 Make the recoveryamp in two stages for example a 12DW7(7247, ECC832) with the 12AX7 section directly after the springs, then the reverb-pot followed by the 12AU7.The return to the mixpoint is done via a highohm resistor 500-700k.

This way you will achieve a loss of less than 3dB.